ooberbot

A Discord LLM bot that runs using oobabooga's text-generation-webui

Be warned, this bot is barebones and probably trash. I'm not great with programming or python in general. You have been warned.

Requirements:

• Python
• oobabooga's text-generation-webui
• Linux, or MacOS
• A brain

Installation Instructions:

1. Open a terminal window and type:

git clone https://github.com/notmugi/ooberbot && cd ooberbot

Or download the project from github manually and place it at your desired location.

2. Next, give the install script execution permissions:

chmod +x install.sh

3. Finally, to run the installer, type:

./install.sh

This should copy your ooberbot config to ~/.config/ooberbot, and then install all the required python dependencies from install.txt using pip.

4. Assuming you already have a discord bot set up on the Discord Developer portal, You will need to copy your bot's Discord Token into the "token" field in ~/.config/ooberbot/config.yaml.

5. Generate an invite link in the dev portal and invite your bot to your desired server. or DM it. up to you.

6. Find textgen_default_cfg and textgen_user_cfg in your config.yaml file. Set these to the corresponding file paths. wherever you decided to install the text-generation-webui.

That's it!

Usage:

At this point, you are free to start using ooberbot.

Simply type python ooberbot.py if you are still in the same folder, or python /path/to/ooberbot.py if you are not.

Currently, the bot is very barebones and will only reply to whatever message it sees, without any rate limiting or anything of that sort.

Commands:

• You may pause the bots interactions with /pause
• you may resume them again with /resume
• You may clear the bots history using /lobotomize

Config:

All config is stored in ~/.config/ooberbot/config.yaml and is commented to explain all features.

Setting "settings_type" in config.yaml to user will use settings.yaml from your text-generation-webui. this is the settings file that gets used when you save your generation settings. otherwise, it uses default which pulls from settings-template.yaml from your text-generation-webui. this is the default config, and should give your bot generic ai responses.

If you'd like to include your own video file, image, text file, gif, audio file, or any other file, for when the bot's history is reset, place it in ~/.config/ooberbot/vids and open config.yaml. Find the line "video_file_name" and specify the filename there. It supports whatever discord will embed. Images, audio, text files, videos, gifs, etc. This behavior is disabled entirely through the included config.yaml file.

Advanced configuration:

If you'd like to add more settings from the webui's config files, feel free to implement more stuff in the settings = {} field. Just follow the current format such as:

"value1": textgencfg['value1'],
"value2": textgencfg['value2'],
"value3": textgencfg['value3']

However, your "preset" contains all the generation information, and "instruction_template" already contains your prompt. I'm not sure what else you may need, but i guess its free for you to edit, so have fun.